Re: [Haskell-cafe] Comment Syntax
Guy guytsalmave...@yahoo.com writes: Out of interest, is there any other language where the comment delimiter is invalid if immediately followed by a symbol? Another quaint example, in shell scripts, lines starting with '#' are comments, except when the first line starts with '#!'. Admittedly, this is still a comment as far as the shell is concerned, it's the OS that is intercepting the comment's contents and acting on it. -k -- If I haven't seen further, it is by standing in the footprints of giants ___ Haskell-Cafe mailing list Haskell-Cafe@haskell.org http://www.haskell.org/mailman/listinfo/haskell-cafe
Re: [Haskell-cafe] Comment Syntax
On 8 June 2011 18:13, Ketil Malde ke...@malde.org wrote: Guy guytsalmave...@yahoo.com writes: Out of interest, is there any other language where the comment delimiter is invalid if immediately followed by a symbol? Another quaint example, in shell scripts, lines starting with '#' are comments, except when the first line starts with '#!'. Admittedly, this is still a comment as far as the shell is concerned, it's the OS that is intercepting the comment's contents and acting on it. And #! in the first line is also treated as a comment in Haskell code so that you can run it as a script. -- Ivan Lazar Miljenovic ivan.miljeno...@gmail.com IvanMiljenovic.wordpress.com ___ Haskell-Cafe mailing list Haskell-Cafe@haskell.org http://www.haskell.org/mailman/listinfo/haskell-cafe
Re: [Haskell-cafe] Comment Syntax
Ivan Lazar Miljenovic ivan.miljeno...@gmail.com writes: And #! in the first line is also treated as a comment in Haskell code so that you can run it as a script. True. But then you're allowed to add arbitrary symbols after it, I think. At least, GHC seems happy about it. -k -- If I haven't seen further, it is by standing in the footprints of giants ___ Haskell-Cafe mailing list Haskell-Cafe@haskell.org http://www.haskell.org/mailman/listinfo/haskell-cafe
Re: [Haskell-cafe] Comment Syntax
On 06/06/2011 22:14, Evan Laforge wrote: Back to Haskell: I agree, the choice of the comment delimiter was not the best in light of the possibility to define operators containing it as a substring. But changing it to have --| start a comment too might break too much code (and eliminating -- as a comment starter would certainly break far too much code). I like that you have to put a space in for haddock comments. I originally posted because I found that --| stood out much more clearly as a structured comment than -- |. ___ Haskell-Cafe mailing list Haskell-Cafe@haskell.org http://www.haskell.org/mailman/listinfo/haskell-cafe
Re: [Haskell-cafe] Comment Syntax
On 7 June 2011 17:41, Guy guytsalmave...@yahoo.com wrote: On 06/06/2011 22:14, Evan Laforge wrote: Back to Haskell: I agree, the choice of the comment delimiter was not the best in light of the possibility to define operators containing it as a substring. But changing it to have --| start a comment too might break too much code (and eliminating -- as a comment starter would certainly break far too much code). I like that you have to put a space in for haddock comments. I originally posted because I found that --| stood out much more clearly as a structured comment than -- |. How does a missing space character make that stand out any more? :/ (Admittedly, I rely more on emacs using a different colour for Haddock comments than non-Haddock comments.) -- Ivan Lazar Miljenovic ivan.miljeno...@gmail.com IvanMiljenovic.wordpress.com ___ Haskell-Cafe mailing list Haskell-Cafe@haskell.org http://www.haskell.org/mailman/listinfo/haskell-cafe
Re: [Haskell-cafe] Comment Syntax
On 07/06/2011 10:45, Ivan Lazar Miljenovic wrote: On 7 June 2011 17:41, Guyguytsalmave...@yahoo.com wrote: On 06/06/2011 22:14, Evan Laforge wrote: Back to Haskell: I agree, the choice of the comment delimiter was not the best in light of the possibility to define operators containing it as a substring. But changing it to have --| start a comment too might break too much code (and eliminating -- as a comment starter would certainly break far too much code). I like that you have to put a space in for haddock comments. I originally posted because I found that --| stood out much more clearly as a structured comment than -- |. How does a missing space character make that stand out any more? :/ (Admittedly, I rely more on emacs using a different colour for Haddock comments than non-Haddock comments.) Try it without emacs :-) ___ Haskell-Cafe mailing list Haskell-Cafe@haskell.org http://www.haskell.org/mailman/listinfo/haskell-cafe
Re: [Haskell-cafe] Comment Syntax
On 7 June 2011 17:50, Guy guytsalmave...@yahoo.com wrote: On 07/06/2011 10:45, Ivan Lazar Miljenovic wrote: On 7 June 2011 17:41, Guyguytsalmave...@yahoo.com wrote: I originally posted because I found that --| stood out much more clearly as a structured comment than -- |. How does a missing space character make that stand out any more? :/ (Admittedly, I rely more on emacs using a different colour for Haddock comments than non-Haddock comments.) Try it without emacs :-) I've also read un-highlighted Haskell code; I don't see --| standing out any more than -- | does. My guess is that you just get used to it... Another argument against special-casing --|: what happens if you want to use a _different_ documentation generator (I don't know why you would, but someone might) than Haddock, which uses a different markup identifier? -- Ivan Lazar Miljenovic ivan.miljeno...@gmail.com IvanMiljenovic.wordpress.com ___ Haskell-Cafe mailing list Haskell-Cafe@haskell.org http://www.haskell.org/mailman/listinfo/haskell-cafe
Re: [Haskell-cafe] Comment Syntax
On 07/06/2011 10:55, Ivan Lazar Miljenovic wrote: On 7 June 2011 17:50, Guyguytsalmave...@yahoo.com wrote: On 07/06/2011 10:45, Ivan Lazar Miljenovic wrote: On 7 June 2011 17:41, Guyguytsalmave...@yahoo.comwrote: I originally posted because I found that --| stood out much more clearly as a structured comment than -- |. How does a missing space character make that stand out any more? :/ (Admittedly, I rely more on emacs using a different colour for Haddock comments than non-Haddock comments.) Try it without emacs :-) I've also read un-highlighted Haskell code; I don't see --| standing out any more than -- | does. My guess is that you just get used to it... Another argument against special-casing --|: what happens if you want to use a _different_ documentation generator (I don't know why you would, but someone might) than Haddock, which uses a different markup identifier? Out of interest, is there any other language where the comment delimiter is invalid if immediately followed by a symbol? Haskell seems to be parsing -- as if it was an operator (hence other legal lexemes could mean something else). Other languages say stop parsing here, so the comment delimiter can be followed by anything. (I could, of course, be completely wrong there.) ___ Haskell-Cafe mailing list Haskell-Cafe@haskell.org http://www.haskell.org/mailman/listinfo/haskell-cafe
Re: [Haskell-cafe] Comment Syntax
Guy guytsalmave...@yahoo.com writes: Out of interest, is there any other language where the comment delimiter is invalid if immediately followed by a symbol? Perl has a rather infamous example where the comment syntax may depend on run-time properties - would that count? whatever / 25 ; # / ; die this dies!; Depending on the definition of 'whatever', the # might introduce a comment - or it might not. Anyway, I think very few languages allow definition of new symbolic operators, so it's hard to make a useful comparison. -k -- If I haven't seen further, it is by standing in the footprints of giants ___ Haskell-Cafe mailing list Haskell-Cafe@haskell.org http://www.haskell.org/mailman/listinfo/haskell-cafe
Re: [Haskell-cafe] Comment Syntax
On 07/06/2011 10:55, Ivan Lazar Miljenovic wrote: Another argument against special-casing --|: what happens if you want to use a _different_ documentation generator (I don't know why you would, but someone might) than Haddock, which uses a different markup identifier? We can declare new operators - why not new comment delimiters? That way Haddock could define --| and --^, other generators could declare whatever they wanted. The downside of this is that Haddock would have to be imported into every file which uses the markup. ___ Haskell-Cafe mailing list Haskell-Cafe@haskell.org http://www.haskell.org/mailman/listinfo/haskell-cafe
Re: [Haskell-cafe] Comment Syntax
On 7/06/2011, at 9:36 PM, Guy wrote: Out of interest, is there any other language where the comment delimiter is invalid if immediately followed by a symbol? Not exactly what you asked, but in some implementations of Algol, comment This is a comment; commentThis is a syntax error; And in the documentation of the M4 macro processor, where changecom(L,R) sets left and right delimiters for comments and changequote(L,R) sets left and right delimiters for quoting, the effect of having one of these left delimiters a prefix of the other is undefined. When it happens by accident the results are, well, confusing. But after doing changecom(dnlx) you find that dnl xyz is a call to the built in macro Discard until New Line, while dnlx yz is a comment. There have been Prolog systems on IBM machines where the set of legal tokens depended on which operators had been declared. For example, if you had declared -- as an operator, then was two -- tokens, but otherwise four - tokens. What they did with comments I shudder to think. ___ Haskell-Cafe mailing list Haskell-Cafe@haskell.org http://www.haskell.org/mailman/listinfo/haskell-cafe
Re: [Haskell-cafe] Comment Syntax
On 4/06/2011, at 5:12 AM, Andrew Coppin wrote: I'm curious to know why anybody thought that -- was a good comment marker in the first place. (I'm curious because Haskell isn't the only language to have made this strange choice.) Indeed. The Wikipedia lists Euphoria, Haskell, SQL, Ada, AppleScript, Eiffel, Lua, and VHDL. Presumably it was chosen for its resemblance to the em dash. SNOBOL had a neat hack. You could begin a comment with any of ⎡⎢⎣ making it possible to write blocks that looked like ⎡ A comment that ⎢ stretches over ⎣ several lines. APL's lamp symbol had the advantage of being wholly new and so having no prior conflicting uses. ⍝ has been in Unicode for a relatively long time. Bearing in mind that the characters that have been used to begin end of line comments include *, /, ;, !, #, %, and $, it's not clear that there's anything _that_ regrettable about -- . ___ Haskell-Cafe mailing list Haskell-Cafe@haskell.org http://www.haskell.org/mailman/listinfo/haskell-cafe
Re: [Haskell-cafe] Comment Syntax
This whole discussion is reminding me of Wadler's Law of Language Design [1], it's nice to see that in 15 years things haven't changed much! WADLER'S LAW OF LANGUAGE DESIGN In any language design, the total time spent discussing a feature in this list is proportional to two raised to the power of its position. 0. Semantics 1. Syntax 2. Lexical syntax 3. Lexical syntax of comments [1] http://www.informatik.uni-kiel.de/~curry,/listarchive/0017.html ___ Haskell-Cafe mailing list Haskell-Cafe@haskell.org http://www.haskell.org/mailman/listinfo/haskell-cafe
Re: [Haskell-cafe] Comment Syntax
Nicolas Wu schrieb: This whole discussion is reminding me of Wadler's Law of Language Design [1], it's nice to see that in 15 years things haven't changed much! WADLER'S LAW OF LANGUAGE DESIGN In any language design, the total time spent discussing a feature in this list is proportional to two raised to the power of its position. 0. Semantics 1. Syntax 2. Lexical syntax 3. Lexical syntax of comments [1] http://www.informatik.uni-kiel.de/~curry,/listarchive/0017.html It's the first time, that I see a discussion about comment syntax in Haskell-Cafe. ___ Haskell-Cafe mailing list Haskell-Cafe@haskell.org http://www.haskell.org/mailman/listinfo/haskell-cafe
Re: [Haskell-cafe] Comment Syntax
Bearing in mind that the characters that have been used to begin end of line comments include *, /, ;, !, #, %, and $, it's not clear that there's anything _that_ regrettable about -- . Recall that the problem is not with isolated characters, but whole strings. -- a is a comment, --a is a comment, but ---a is not. That is done wrong. Either make them all comments, or make them none comments. {- a -} is a comment, {-a-} is a comment, {--a--} is a comment, {---} is a comment. That is done right. Similarly, in C, /***/ is a comment; in C++, /// is a comment (compare with Haskell's ---!); in LaTeX, %%@#$^* is a comment. These are all done right. Consistently. ___ Haskell-Cafe mailing list Haskell-Cafe@haskell.org http://www.haskell.org/mailman/listinfo/haskell-cafe
Re: [Haskell-cafe] Comment Syntax
On Montag, 6. Juni 2011, 19:08, Albert Y. C. Lai wrote: Bearing in mind that the characters that have been used to begin end of line comments include *, /, ;, !, #, %, and $, it's not clear that there's anything _that_ regrettable about -- . Recall that the problem is not with isolated characters, but whole strings. -- a is a comment, --a is a comment, but ---a is not. It is. Report, section 2.3: 'An ordinary comment begins with a sequence of two or more consecutive dashes (e.g. --) and extends to the following newline. The sequence of dashes must not form part of a legal lexeme. For example, “--” or “|--” do not begin a comment, because both of these are legal lexemes; however “-- foo” does start a comment.' A sequence of two or more dashes does not begin an end-of-line-comment if and only if it is part of a lexeme containing only symbols and at least one non-dash symbol. It might be a good idea to include whitespace in the comment delimiter: end-of-line-comment is begun by whitespace--{-}whitespace. ___ Haskell-Cafe mailing list Haskell-Cafe@haskell.org http://www.haskell.org/mailman/listinfo/haskell-cafe
Re: [Haskell-cafe] Comment Syntax
On 2011-06-06 13:08 -0400, Albert Y. C. Lai wrote: Recall that the problem is not with isolated characters, but whole strings. [...] in LaTeX, %%@#$^* is a comment. This example probably does not help your position. Since (La)TeX allows the comment character to be changed at any time, the above is not necessarily a comment. Furthermore, even with the default character classifications, \% does not introduce a comment. \% not introducing a comment in (La)TeX doesn't seem a whole lot different from --- not introducing a comment in Haskell. -- Nick Bowler, Elliptic Technologies (http://www.elliptictech.com/) ___ Haskell-Cafe mailing list Haskell-Cafe@haskell.org http://www.haskell.org/mailman/listinfo/haskell-cafe
Re: [Haskell-cafe] Comment Syntax
On 2011-06-06 13:39 -0400, Nick Bowler wrote: On 2011-06-06 13:08 -0400, Albert Y. C. Lai wrote: Recall that the problem is not with isolated characters, but whole strings. [...] in LaTeX, %%@#$^* is a comment. This example probably does not help your position. Since (La)TeX allows the comment character to be changed at any time, the above is not necessarily a comment. Furthermore, even with the default character classifications, \% does not introduce a comment. \% not introducing a comment in (La)TeX doesn't seem a whole lot different from --- not introducing a comment in Haskell. And as was pointed out elsethread, --- /does/ in fact introduce a comment in Haskell. So the above should read: \% ... doesn't seem a whole lot different from --| ... -- Nick Bowler, Elliptic Technologies (http://www.elliptictech.com/) ___ Haskell-Cafe mailing list Haskell-Cafe@haskell.org http://www.haskell.org/mailman/listinfo/haskell-cafe
Re: [Haskell-cafe] Comment Syntax
On 11-06-06 01:34 PM, Daniel Fischer wrote: On Montag, 6. Juni 2011, 19:08, Albert Y. C. Lai wrote: Recall that the problem is not with isolated characters, but whole strings. -- a is a comment, --a is a comment, but ---a is not. It is. Report, section 2.3: Sorry. Then --| is not a comment. In C++, //| is a comment (compare with Haskell's --|!). ___ Haskell-Cafe mailing list Haskell-Cafe@haskell.org http://www.haskell.org/mailman/listinfo/haskell-cafe
Re: [Haskell-cafe] Comment Syntax
On Monday 06 June 2011, 19:51:44, Albert Y. C. Lai wrote: On 11-06-06 01:34 PM, Daniel Fischer wrote: On Montag, 6. Juni 2011, 19:08, Albert Y. C. Lai wrote: Recall that the problem is not with isolated characters, but whole strings. -- a is a comment, --a is a comment, but ---a is not. It is. Report, section 2.3: Sorry. Then --| is not a comment. In C++, //| is a comment (compare with Haskell's --|!). True. But then, to my knowledge, C++ has a fixed small set of operators, so the problem of a comment delimiter possibly being part of a different lexeme doesn't arise there. There is, however, a somewhat similar problem in C and C++: z = *x/*y; // oops */ Back to Haskell: I agree, the choice of the comment delimiter was not the best in light of the possibility to define operators containing it as a substring. But changing it to have --| start a comment too might break too much code (and eliminating -- as a comment starter would certainly break far too much code). ___ Haskell-Cafe mailing list Haskell-Cafe@haskell.org http://www.haskell.org/mailman/listinfo/haskell-cafe
Re: [Haskell-cafe] Comment Syntax
Back to Haskell: I agree, the choice of the comment delimiter was not the best in light of the possibility to define operators containing it as a substring. But changing it to have --| start a comment too might break too much code (and eliminating -- as a comment starter would certainly break far too much code). I like that you have to put a space in for haddock comments. ___ Haskell-Cafe mailing list Haskell-Cafe@haskell.org http://www.haskell.org/mailman/listinfo/haskell-cafe
Re: [Haskell-cafe] Comment Syntax
Albert Y. C. Lai tre...@vex.net writes: On 11-06-04 02:20 AM, Roman Cheplyaka wrote: It is, for my taste, a good comment marker, because of its resemblance to a dash. It makes the code look like real text: let y = x + 1 -- increment x COBOL is real text, if that is what you want. MOVE PROGRAMMER TO SARCASM -- Jón Fairbairn jon.fairba...@cl.cam.ac.uk ___ Haskell-Cafe mailing list Haskell-Cafe@haskell.org http://www.haskell.org/mailman/listinfo/haskell-cafe
Re: [Haskell-cafe] Comment Syntax
* Andrew Coppin andrewcop...@btinternet.com [2011-06-03 18:12:04+0100] On 03/06/2011 05:02 PM, Albert Y. C. Lai wrote: I propose that only {- -} is comment; that is, -- is an operator token and not a marker of comments. I'm curious to know why anybody thought that -- was a good comment marker in the first place. (I'm curious because Haskell isn't the only language to have made this strange choice.) It is, for my taste, a good comment marker, because of its resemblance to a dash. It makes the code look like real text: let y = x + 1 -- increment x -- Roman I. Cheplyaka :: http://ro-che.info/ ___ Haskell-Cafe mailing list Haskell-Cafe@haskell.org http://www.haskell.org/mailman/listinfo/haskell-cafe
Re: [Haskell-cafe] Comment Syntax
Roman Cheplyaka r...@ro-che.info writes: * Andrew Coppin andrewcop...@btinternet.com [2011-06-03 18:12:04+0100] On 03/06/2011 05:02 PM, Albert Y. C. Lai wrote: I propose that only {- -} is comment; that is, -- is an operator token and not a marker of comments. I'm curious to know why anybody thought that -- was a good comment marker in the first place. (I'm curious because Haskell isn't the only language to have made this strange choice.) It is, for my taste, a good comment marker, because of its resemblance to a dash. It makes the code look like real text: let y = x + 1 -- increment x And when the language was first defined, there was no real dash available for this purpose. Nowadays we could use the unicode em dash (U+2014, — if it survices nntp), and free up -- for other purposes. Of course, its hard to predict what effect this would have when people didn't know how to enter it, and although it's quite distinctive in proportional fonts, the difference between - and — in monospace fonts is much harder to see. -- Jón Fairbairn jon.fairba...@cl.cam.ac.uk http://www.chaos.org.uk/~jf/Stuff-I-dont-want.html (updated 2010-09-14) ___ Haskell-Cafe mailing list Haskell-Cafe@haskell.org http://www.haskell.org/mailman/listinfo/haskell-cafe
Re: [Haskell-cafe] Comment Syntax
On 11-06-04 02:20 AM, Roman Cheplyaka wrote: It is, for my taste, a good comment marker, because of its resemblance to a dash. It makes the code look like real text: let y = x + 1 -- increment x COBOL is real text, if that is what you want. ___ Haskell-Cafe mailing list Haskell-Cafe@haskell.org http://www.haskell.org/mailman/listinfo/haskell-cafe
Re: [Haskell-cafe] Comment Syntax
*Touché.* Nice one. 2011/6/4 Albert Y. C. Lai tre...@vex.net On 11-06-04 02:20 AM, Roman Cheplyaka wrote: It is, for my taste, a good comment marker, because of its resemblance to a dash. It makes the code look like real text: let y = x + 1 -- increment x COBOL is real text, if that is what you want. ___ Haskell-Cafe mailing list Haskell-Cafe@haskell.org http://www.haskell.org/mailman/listinfo/haskell-cafe ___ Haskell-Cafe mailing list Haskell-Cafe@haskell.org http://www.haskell.org/mailman/listinfo/haskell-cafe
[Haskell-cafe] Comment Syntax
-- followed by a symbol does not start a comment, thus for example, haddock declarations must begin with -- |, and not --|. What might --| mean, if not a comment? It doesn't seem possible to define it as an operator. ___ Haskell-Cafe mailing list Haskell-Cafe@haskell.org http://www.haskell.org/mailman/listinfo/haskell-cafe
Re: [Haskell-cafe] Comment Syntax
On Fri, Jun 3, 2011 at 10:32, Guy guytsalmave...@yahoo.com wrote: -- followed by a symbol does not start a comment, thus for example, haddock declarations must begin with -- |, and not --|. What might --| mean, if not a comment? It doesn't seem possible to define it as an operator. GHCi, at least, allows it. Prelude let (--|) = (+) Prelude 1 --| 2 3 --Max ___ Haskell-Cafe mailing list Haskell-Cafe@haskell.org http://www.haskell.org/mailman/listinfo/haskell-cafe
Re: [Haskell-cafe] Comment Syntax
On 3 June 2011 18:32, Guy guytsalmave...@yahoo.com wrote: -- followed by a symbol does not start a comment, thus for example, haddock declarations must begin with -- |, and not --|. What might --| mean, if not a comment? It doesn't seem possible to define it as an operator. Sure you can; --| is a valid operator. -- Ivan Lazar Miljenovic ivan.miljeno...@gmail.com IvanMiljenovic.wordpress.com ___ Haskell-Cafe mailing list Haskell-Cafe@haskell.org http://www.haskell.org/mailman/listinfo/haskell-cafe
Re: [Haskell-cafe] Comment Syntax
-- followed by a symbol does not start a comment, thus for example, haddock declarations must begin with -- |, and not --|. What might --| mean, if not a comment? It doesn't seem possible to define it as an operator. GHCi, at least, allows it. Prelude let (--|) = (+) Prelude 1 --| 2 3 I believe the motivating example that persuaded the Language Committee to allow these symbols was -- which is not of course used anywhere in the standard libraries, but is an extremely nice symbol to have available in user code. Regards, Malcolm ___ Haskell-Cafe mailing list Haskell-Cafe@haskell.org http://www.haskell.org/mailman/listinfo/haskell-cafe
Re: [Haskell-cafe] Comment Syntax
On 03/06/2011 12:01, Malcolm Wallace wrote: I believe the motivating example that persuaded the Language Committee to allow these symbols was -- which is not of course used anywhere in the standard libraries, but is an extremely nice symbol to have available in user code. Seeing as no library actually defines such a symbol, is it worth forcing an extra space into comments? --| and --^ would be an extremely nice symbols for haddock comments. ___ Haskell-Cafe mailing list Haskell-Cafe@haskell.org http://www.haskell.org/mailman/listinfo/haskell-cafe
Re: [Haskell-cafe] Comment Syntax
On 3 June 2011 19:19, Guy guytsalmave...@yahoo.com wrote: On 03/06/2011 12:01, Malcolm Wallace wrote: I believe the motivating example that persuaded the Language Committee to allow these symbols was -- which is not of course used anywhere in the standard libraries, but is an extremely nice symbol to have available in user code. Seeing as no library actually defines such a symbol, is it worth forcing an extra space into comments? --| and --^ would be an extremely nice symbols for haddock comments. We already have to forms of comments: -- and {- -}; why do we need another two? Think of it this way: the -- defines the comment, and then we use the | or ^ to indicate to _haddock_ to interpret this differently. I don't think we need to special-case symbols starting with -- just to avoid having to put a space in to delimit the comment indicator and the markup indicator. That said, I have a use for -- as a symbol... if it wasn't that I'd also want something like -- to follow suit, and as such it'll just be easier to use something like .-. and .--. (for creating directed and undirected edges respectively in graphs). -- Ivan Lazar Miljenovic ivan.miljeno...@gmail.com IvanMiljenovic.wordpress.com ___ Haskell-Cafe mailing list Haskell-Cafe@haskell.org http://www.haskell.org/mailman/listinfo/haskell-cafe
Re: [Haskell-cafe] Comment Syntax
On 03/06/2011 12:26, Ivan Lazar Miljenovic wrote: On 3 June 2011 19:19, Guyguytsalmave...@yahoo.com wrote: On 03/06/2011 12:01, Malcolm Wallace wrote: I believe the motivating example that persuaded the Language Committee to allow these symbols was -- which is not of course used anywhere in the standard libraries, but is an extremely nice symbol to have available in user code. Seeing as no library actually defines such a symbol, is it worth forcing an extra space into comments? --| and --^ would be an extremely nice symbols for haddock comments. We already have to forms of comments: -- and {- -}; why do we need another two? Think of it this way: the -- defines the comment, and then we use the | or ^ to indicate to _haddock_ to interpret this differently. I don't think we need to special-case symbols starting with -- just to avoid having to put a space in to delimit the comment indicator and the markup indicator. I wasn't proposing additional comment symbols; I'm proposing that anything beginning with -- is a comment. ___ Haskell-Cafe mailing list Haskell-Cafe@haskell.org http://www.haskell.org/mailman/listinfo/haskell-cafe
Re: [Haskell-cafe] Comment Syntax
2011/6/3 Guy guytsalmave...@yahoo.com: I wasn't proposing additional comment symbols; I'm proposing that anything beginning with -- is a comment. I use -- as a infix operator to describe types in Template Haskell. So I too oppose your proposal. ;) ___ Haskell-Cafe mailing list Haskell-Cafe@haskell.org http://www.haskell.org/mailman/listinfo/haskell-cafe
Re: [Haskell-cafe] Comment Syntax
On Fri, 03 Jun 2011 12:19:31 +0300, Guy guytsalmave...@yahoo.com wrote: On 03/06/2011 12:01, Malcolm Wallace wrote: I believe the motivating example that persuaded the Language Committee to allow these symbols was -- which is not of course used anywhere in the standard libraries, but is an extremely nice symbol to have available in user code. Seeing as no library actually defines such a symbol, is it worth forcing an extra space into comments? [..] That's not true, xmonad[0] for example defines a ---operator; and I would find making exceptions for --| and --^ very inconsistent and annoying. [0] http://xmonad.org/xmonad-docs/xmonad/XMonad-ManageHook.html#v%3A--%3E pgp8FyrB33uh7.pgp Description: PGP signature ___ Haskell-Cafe mailing list Haskell-Cafe@haskell.org http://www.haskell.org/mailman/listinfo/haskell-cafe
Re: [Haskell-cafe] Comment Syntax
On 3 June 2011 20:32, Daniel Schoepe daniel.scho...@googlemail.com wrote: On Fri, 03 Jun 2011 12:19:31 +0300, Guy guytsalmave...@yahoo.com wrote: On 03/06/2011 12:01, Malcolm Wallace wrote: I believe the motivating example that persuaded the Language Committee to allow these symbols was -- which is not of course used anywhere in the standard libraries, but is an extremely nice symbol to have available in user code. Seeing as no library actually defines such a symbol, is it worth forcing an extra space into comments? [..] That's not true, xmonad[0] for example defines a ---operator; and I would find making exceptions for --| and --^ very inconsistent and annoying. [0] http://xmonad.org/xmonad-docs/xmonad/XMonad-ManageHook.html#v%3A--%3E I _knew_ I had seen -- somewhere... *glares at Hayoo for not returning said results*. -- Ivan Lazar Miljenovic ivan.miljeno...@gmail.com IvanMiljenovic.wordpress.com ___ Haskell-Cafe mailing list Haskell-Cafe@haskell.org http://www.haskell.org/mailman/listinfo/haskell-cafe
Re: [Haskell-cafe] Comment Syntax
Am 03.06.2011 10:32, schrieb Guy: What might --| mean, if not a comment? It doesn't seem possible to define it as an operator. Obviously, anyone who is going to write a formal logic framework would want to define the following operators ;) : T |- phi: T proves phi T |-- phi: T proves phi directly (by application of a single rule) phi -| T: phi is proven by T phi --| T: phi is proven by T directly ___ Haskell-Cafe mailing list Haskell-Cafe@haskell.org http://www.haskell.org/mailman/listinfo/haskell-cafe
Re: [Haskell-cafe] Comment Syntax
On Fri, Jun 3, 2011 at 05:19, Guy guytsalmave...@yahoo.com wrote: On 03/06/2011 12:01, Malcolm Wallace wrote: I believe the motivating example that persuaded the Language Committee to allow these symbols was -- which is not of course used anywhere in the standard libraries, but is an extremely nice symbol to have available in user code. Seeing as no library actually defines such a symbol, is it worth forcing an extra space into comments? --| and --^ would be an extremely nice symbols for haddock comments. http://xmonad.org/xmonad-docs/xmonad/XMonad-ManageHook.html#v%3A--%3E ___ Haskell-Cafe mailing list Haskell-Cafe@haskell.org http://www.haskell.org/mailman/listinfo/haskell-cafe
Re: [Haskell-cafe] Comment Syntax
I propose that only {- -} is comment; that is, -- is an operator token and not a marker of comments. Two birds in one stone: 1. Removes the cause of the mistake of writing a haddock comment as --| That is, if no one writes any comment with -- then no one writes any haddock comment with --| --^ etc. 2. Opens up the option of using -- for prefix unary negation, so that - is unambiguously infix subtraction and therefore (- x) is consistently right sectioning of infix subtraction. ___ Haskell-Cafe mailing list Haskell-Cafe@haskell.org http://www.haskell.org/mailman/listinfo/haskell-cafe
Re: [Haskell-cafe] Comment Syntax
On 03/06/2011 05:02 PM, Albert Y. C. Lai wrote: I propose that only {- -} is comment; that is, -- is an operator token and not a marker of comments. I'm curious to know why anybody thought that -- was a good comment marker in the first place. (I'm curious because Haskell isn't the only language to have made this strange choice.) I once tried defining a ++ operator and then when I went to define a matching -- operator... ___ Haskell-Cafe mailing list Haskell-Cafe@haskell.org http://www.haskell.org/mailman/listinfo/haskell-cafe
Re: [Haskell-cafe] Comment Syntax
On 4 June 2011 02:02, Albert Y. C. Lai tre...@vex.net wrote: I propose that only {- -} is comment; that is, -- is an operator token and not a marker of comments. Two birds in one stone: 1. Removes the cause of the mistake of writing a haddock comment as --| That is, if no one writes any comment with -- then no one writes any haddock comment with --| --^ etc. 2. Opens up the option of using -- for prefix unary negation, so that - is unambiguously infix subtraction and therefore (- x) is consistently right sectioning of infix subtraction. I think it's probably too late to make such a sweeping change; besides, having some kind of inline comment indicator (which goes to the end of line) is useful. -- Ivan Lazar Miljenovic ivan.miljeno...@gmail.com IvanMiljenovic.wordpress.com ___ Haskell-Cafe mailing list Haskell-Cafe@haskell.org http://www.haskell.org/mailman/listinfo/haskell-cafe